SUMMARY: Suspect Seagate Barracuda

From: Chris Wozniak TISC (chris@tisc.edu.au)
Date: Mon Mar 20 1995 - 01:27:23 CST


I have received 6 answers to my query about suspect (possibly)
Seagate Barracuda drive, producing SCSI error messages at random.

Three of those answers were suggesting either cable length, terminator,
or a plug problem. I've checked and excluded all these possibilities,
but the message was still there.
Alex Barclay, Rainer Ullrich and Tom Crummey have all pointed me in the
right direction - Seagate firmware/microcode revision.
I've arranged with the supplier its replacement during the quiet period
after Easter.

----------------------------------------------------------------------
>Hello,
>
>Which firmware revision have you on the disk? I had the same sort of trouble
>on my Sparc 20 with this disk. That was with forware revision 0014. With 0017
>I have no trouble.
>
>You can check the revision in format:
>
>< Seagate Barracuda ST15150N xxxx>
>
>xxxx is the revision number.
>
>If you have lower the 0017 ask your vendor/Seagate for a replacement.
>
>Hope this helps.
>
>Of course it could be the usual cable too long/passive terminators/too many
>devices problems that dog fast SCSI disks on SUN controllers.
>
>--
>Tom.
>
----------------------------------------------------------------------
>
>Hi Chris,
>
>we have had some equal problems with the drive. The solution was to download
>an new microcode. The original revision was 10. The bug fixed revision is 12.
>But there is one big problem. The download-program from seagate is only
>available for an apple mac. Today, I'll "hack" a download-program for sun. For
>this I'll need a while. To get the revision, you can use the probe-scsi-command
>or scsiinfo.
>
>If it's necessary, I can send you the microcode. When I'am finished the
>program, I'am able to send you the tested program, if you need.
>Please tell me.
>
>Regards
>Rainer
----------------------------------------------------------------------

>Chris,
>
>I had exactly the same problem with a 4GB barracuda drive. It happens under
>high random access load ( in my case = large Oracle 7 DB ).
>
>The problem occurs with SCSI-II tag commands. SUN and Seagate have a very
>different view of the world. Neither will accept the blame but both offer
>fixes.
>
>There are two cures that I know about.....
>
>1) Have the drive firmware upgraded. Seagate did this for me for free. This
>definately
>worked
>
>2) Upgrade your version of Solaris. ( are you running SunOS still?? )
> When I had the problems I was running Solaris 2.2, the problem also
> manifested itself under SunOS 4.1.3 and Solaris 2.3.
> I am now running Solaris 2.4 but performed this upgrade once the drive
>was
> fixed and so don't know if 2.4 fixes the problem.
>
>All in all I would _reccomend_ the drive firmware upgrade. This is a
>non-destructive operation that takes less than 20 minutes.
>
>Good luck,
>
>Alex.

Thanks a lot to the people who replied:

 Glenn.Satchell@uniq.com.au
 Ross.Stocks.INSDRS01@nt.com
 Kevin.Sheehan@uniq.com.au
 Mr T Crummey (DIJ) <tom@sees.bangor.ac.uk>
 Rainer.Ullrich@teefax.mch.sni.de
 Barclay, Alex <alex@mlbg79.btlabsmh.bt.co.uk>

-- 
Chris A. Wozniak
TISC Inc.

ph. +61 9 386 8236 e-mail chris@tisc.edu.au fax +61 9 386 6589



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:10:19 CDT